Association of DFNA5,SYK,and NELL1 variants along with HPV infection in oral cancer among the prolonged tobacco-chewers

Abstract

Southeast Asia, especially India, is well known for the highest use of smokeless tobacco. These products are known to induce oral squamous cell carcinoma. However, not all long-term tobacco-chewers develop oral squamous cell carcinoma. In addition, germline variants play a crucial role in susceptibility, prognosis, development, and progression of the disease. These prompted us to study the genetic susceptibility to oral squamous cell carcinoma among the long-term tobacco-chewers. Here, we presented a retrospective study on prolonged tobacco-chewers of Northeast India to identify the potential protective or risk-associated germline variants in tobacco-related oral squamous cell carcinoma along with HPV infection. Targeted re-sequencing (n = 60) of 170 genetic regions from 75 genes was carried out in Ion-PGM™ and validation (n = 116) of the observed variants was done using Sequenom iPLEX MassARRAY™ platform followed by polymerase chain reaction–based HPV genotyping and p16-immunohistochemistry study. Subsequently, estimation of population structure, different statistical and in silico approaches were undertaken. We identified one nonsense-mediated mRNA decay transcript variant in the DFNA5 region (rs2237306), associated with Benzo(a)pyrene, as a protective factor (odds ratio = 0.33; p = 0.009) and four harmful (odds ratio > 2.5; p < 0.05) intronic variants, rs182361, rs290974, and rs169724 in SYK and rs1670661 in NELL1 region, involved in genetic susceptibility to tobacco- and HPV-mediated oral oncogenesis. Among the oral squamous cell carcinoma patients, 12.6% (11/87) were HPV positive, out of which 45.5% (5/11) were HPV16-infected, 27.3% (3/11) were HPV18-infected, and 27.3% (3/11) had an infection of both subtypes. Multifactor dimensionality reduction analysis showed that the interactions among HPV and NELL1 variant rs1670661 with age and gender augmented the risk of both non-tobacco- and tobacco-related oral squamous cell carcinoma, respectively. These suggest that HPV infection may be one of the important risk factors for oral squamous cell carcinoma in this population. Finally, we newly report a DFNA5 variant probably conferring protection via nonsense-mediated mRNA decay pathway against tobacco-related oral squamous cell carcinoma. Thus, the analytical approach used here can be useful in predicting the population-specific significant variants associated with oral squamous cell carcinoma in any heterogeneous population.

Keywords

Smokeless tobacco prolonged tobacco-chewers heterogeneous populations tobacco-related oral squamous cell carcinoma germline variant protective single nucleotide polymorphism nonsense-mediated mRNA decay pathway HPV-mediated oncogenesis

Introduction

There is a wide geographical variation in the incidence of oral squamous cell carcinoma (OSCC) with the highest rate found in the Indian subcontinent (India, Pakistan, Sri Lanka, Bangladesh, etc.) contributing to 25% of the new incident cases.¹ Among these countries, approximately 1 million OSCC cases are being diagnosed each year in India, with a male-to-female ratio of 2:1 and the most common site is gingivobuccal complex followed by the tongue.² Among the total oral cancer cases across the world, nearly 30% occur in India, whereas the western world and USA have less than 10% and 3% oral cancer cases, respectively.³

In India, oral cancer is ranked as the most common cancer types in males, with an annual incidence of 53,842 and is the fifth most common cancer in females, with a yearly incidence of 23,161.⁴ Generally, 90%–95% of all the oral cancer cases are found to be squamous cell carcinomas (SCC).^5,6 The wide discrepancy in the incidence, as well as mortality rates of oral cancer in different parts around the world, is due to the combined effects of the regional differences in the prevalence of disease-specific risk factors, aging of the population, and other genetic predisposing factors.^3,7

Some well-established risk factors such as tobacco, betel quid chewing, smoking, alcoholism, human papilloma virus (HPV) infections,^5,6,8,9 and poor oral hygiene⁵ are significantly associated with this anatomical and clinical subtype of head and neck cancer. Among these factors, smokeless tobacco (SLT) is one of the single most important causes for the higher incidence of oral cancer.¹⁰ Globally, almost 90% of OSCC cases are associated with tobacco habits in different forms.¹¹ It has been reported that nearly 85% of the total SLT usage burden is in Southeast Asia.¹² A recent data from national surveys showed a clear shift in the product preference from smoking to SLT in Southeast Asia.¹³ Subsequently, cancers of the buccal mucosa, tongue along with hypo-pharynx, are very common, especially in India,¹⁴ attributing to 74% of the total global SLT usage burden.¹²

According to the Global Adult Tobacco Survey (GATS) of the Ministry of Health and Family Welfare released in June 2017, nearly 199 million Indians were using SLT.¹⁵ According to the consolidated population-based cancer registry (PBCR) report published by the Indian Council of Medical Research (ICMR) through their National Cancer Registry Program (NCRP) 2016, it has been projected that by 2020, tobacco-related oral cancer will contribute to 30% of the total number of incidences (i.e. approximately 523,471 cases annually) among all the other anatomical sites.¹⁶ Interestingly, the consumption of SLT products in Northeast (NE) India is significantly higher as compared to the other states of India.¹⁷ Therefore, the incidence of tobacco-related oral cancer in the northeastern part is about 33%.¹⁸ These statistical standpoints make this disease a significant public health importance to India, especially for the NE India, which additionally emphasizes the relevance of this study undertaken.

Apart from tobacco, HPV infection is also an important etiological factor in oral cancer. Approximately 26% of all oral cancer cases had high-risk (HR) type of HPV globally, and HPV16 was the major HR subtype, followed by HPV18.¹⁹ However, this trend is population-specific and also varies with the method undertaken for detecting the viral load.^9,20 The role of this virus in carcinogenesis typically involves the pE7 and pE6 proteins that hamper the p53-, p21-, and pRb-based signaling cascades.²¹

In addition to tobacco and HPV infection, hereditary predisposing genetic factors are very important for the etiopathogenesis of OSCC. The genetic susceptibility is influenced by the germline single nucleotide polymorphisms (SNPs) located within the promoter or other regulatory regions of its genes and is thought to be associated with disease development and progression. Previously, a genome-wide association study (GWAS) conducted on Western Indian population identified 93 SNPs with significantly increased minor allele frequency (MAF) in the tobacco chewing OSCC patients.²² There are also other high-throughput studies, identifying the important polymorphic genes related to this disease globally as well as in other parts of the Indian subcontinent.^23,24 Nevertheless, very scanty or no information about the gene–gene or gene–environment interactions associated with tobacco-related OSCC in NE Indian population is available till date. Therefore, this study was aimed to identify the potential germline variants (protective or risk) associated with tobacco-related oral cancer for the prolonged tobacco-chewers from this population and additionally, we have tailored our analytical approach to derive findings suitable to any heterogeneous population.

In this study, we performed next generation sequencing (NGS)-based targeted re-sequencing, followed by validations using a mass spectroscopy-based high-throughput genotyping technique to identify potential SNPs associated with OSCC in the prolonged tobacco-chewers of NE India. We also performed the population structure estimation, polymerase chain reaction (PCR)-based HPV genotyping, and p16-immunohistochemistry study. Furthermore, multinomial logistic regression (MLR), multifactor dimensionality reduction (MDR) analysis, and different in silico approaches were undertaken to evaluate the effect of the identified SNPs as well as the gene–gene and gene–environment interactions.

Materials and methods

Sampling and demographic data collection

The institutional review board (IRB) of Assam University, Silchar (IEC/AUS/2015-004 dated 4 September 2015), and Cachar Cancer Hospital & Research Centre, Silchar, approved the sampling procedure from different regions of NE India with signed consent form and questionnaire. Data regarding age, gender, type, and duration of consuming smokeless/chewing tobacco of patients as well as healthy subjects were abstracted from hospital records as well as personal interviews.

In total, 50 OSCC patients (case 1) and 74 healthy individuals (without the family history of cancer and without any pre-cancerous lesion/condition) having the prolonged (≥10 years) tobacco-chewing habits along with 52 non-chewers OSCC patients (case 2) from 12 heterogeneous ethnic groups of NE India were considered for this study (see Supplementary Table S1). Surgically removed tissue (7–10 mm) and buccal swab samples were collected aseptically from the histopathologically confirmed OSCC patients and healthy individuals, respectively. In addition, 5-mL peripheral blood was collected for the genetic susceptibility study. The samples were collected during the period of 2013–2015.

Genomic DNA extraction

The genomic DNA from the collected tissue and buccal swab samples was extracted using the standard phenol/chloroform extraction method.²⁵ Genomic DNA from peripheral blood was extracted using QIAamp DNA Blood Mini Kit following manufacturer’s instructions (Qiagen, Hilden, Germany). The extracted DNA samples were quantitated by Qubit® 2.0 Fluorometer (Life Technologies, Grand Island, NY). The purified DNA samples with good quality were used for NGS-based targeted re-sequencing and high-throughput SNP genotyping experiments.

Targeted re-sequencing in the Ion PGM^TM platform on the discovery set

The GWAS related to oral cancer have identified 170 target regions from 75 genes significantly associated with the disease globally and in other parts of Indian subcontinent.^22–24 Two pools of multiplex primers were designed by AmpliSeq™ Designer, a web-based tool by Thermo Fisher Scientific, USA, to cover these 170 targets (see Supplementary Table S2). Total panel size was 31.34 Kb. The targeted re-sequencing experiment was carried out in high-throughput Ion PGM^TM sequencer for 60 case-control samples comprising 20 case-1, 20 healthy, and 20 case-2 individuals as discovery set. A depth of 423.68 ± 17.01 X (mean ± standard error (SE)) with 95.85% uniformity was achieved using the Ion 318^TM chip.

Validation and replication in Sequenom iPLEX MassARRAY^TM platform on the confirmation set

The result obtained from the targeted re-sequencing was validated and then replicated in another set of 116 case-control samples (30 case-1 patients, 54 healthy, and 32 case-2 individuals as confirmation set) using the Sequenom iPLEX MassARRAY^TM platform. For the genotyping of 115 SNPs (see Supplementary Table S3), four pools of primer pairs (see Supplementary Table S4) were designed. Totally, five sets of experiments were performed using these pools for getting the final SNP genotyping results.

PCR-based HPV genotyping

PCR amplification for the detection of HR-HPV types was carried out with the consensus primers GP5 +/GP6 + followed by the primers for genotyping of HPV16 and HPV18.^26,27 The methodology adopted in this study was as per our previous study.⁸

Immunohistochemistry for p16 expression

The surgically removed tissue samples from both the case groups were processed for paraffin block preparation and serial 4-µm sections were consecutively cut from the specimens. The sections were then deparaffinized in xylene and was rehydrated by passing through graded alcohols. Finally, the sections were processed for immunohistochemistry study of P16 antibody (JC-8; Santa Cruz Biotechnology) for cases, negative control (BSA), and positive (cervical cancer) control as instruction given in the kit (DAKO). The representative image is given in Supplementary Figure S1.

In each field, 100 cells were counted and area was marked using measure tool. Similarly, total of 1000 cells were counted and % positivity was calculated. All the counting was done at the objective magnification of 40X. High nuclear and high cytoplasmic staining in greater than 70% of tumor cells was considered positive as suggested by the American Society of Clinical Pathology and College of American pathologists previously.^28,29 The samples were considered as HPV positive only when found positive in both the methods.

Statistical analysis

Genotypes, allele frequencies, and linkage disequilibrium were evaluated using Haploview software³⁰ followed by the stratified association study using PLINK.³¹ The population structure was then estimated by STRUCTURE using StrAuto.^32,33 The number of clusters (K) was estimated by the Delta-K method.³⁴ We calculated odds ratio (OR) and 95% confidence intervals (CIs) using the MLR model. To strengthen the association between variant genotypes and OSCC risk factors, the ORs were calculated after adjusting the confounding factors such as population/ethnicity estimated from the previous step. The statistical analyses were carried out in SPSS software (PASW Statistics 18, Release 18.0.0, 30 July 2009). Pearson’s chi-square test was performed to evaluate p-values and the p-value of <0.05 was considered as statistically significant.

MDR approach

The MDR approach is used to detect high-order gene–gene and gene–environment interactions associated with the disease risk.³⁵ The advantage of this approach over conventional parametric statistical methods is to overcome sample size limitations that often encountered in the retrospective study.³⁶ In this study, we used MDR software package, MDR v3.0.2 (http://multifactordimensionalityreduction.org/) to predict the multifactor models associated with both the types of OSCC susceptibility. The best predictive models were picked out on the basis of the maximum testing balance accuracy (TBA) and corresponding cross-validation consistency (CVC). The p-value of <0.05 was considered to be statistically significant.

Interaction entropy graph

The interaction entropy graph was made based on the MDR result, for better understanding and visualization of interactions among the gene-environmental factors³⁷ as described in our previous studies.^38–40 The entropy graph was constructed to determine the types of interactions (synergistic or redundant) among various genetic as well as environmental factors and then their interaction effects could be compared with the independent effect of each variable.⁴¹

In silico analysis

In silico functional analysis was performed to understand the biological importance of identified significant SNPs. We used Ensembl Variant Effect Predictor tool,⁴² ESRsearch tool,⁴³ and F-SNP⁴⁴ for this. The toxicogenomic analysis was done on the Comparative Toxicogenomics Database.⁴⁵ Subsequently, BiNGO v3.0.3 and Enrichment Map v2.1.0 were also used to get the over-represented Gene Ontology (GO) categories associated with our identified genes.^46,47 Finally, the pathway analysis⁴⁸ by constructing a gene–gene network in GeneMANIA v3.5.0 was done to visualize other genes interacting via some important pathways.⁴⁹

Results

Demographic characteristics of the participants: females are more susceptible to tobacco-related OSCC than males

The summary statistic of patient demography is given in Table 1. The mean age (±SE) of our studied individuals was 49.14 (± 1.22) years. Majority of the OSCC patients (72.5%; 74/102) were in the age group of >49 years, whereas most of the healthy individuals (73%; 54/74) had the age of ≤49 years. Among the female subjects, we observed that 61.1% (33/54) were prolonged (>10 years) tobacco-chewers, out of which 75.8% (25/33) were OSCC patients, whereas, among the males, 74.6% (91/121) were prolonged (>10 years) tobacco-chewers, out of which only 27.5% (25/91) had OSCC. Among the different cancer sites, the buccal mucosa was found to be the most common site (81.4%; 83/102), followed by the tongue (18.6%; 19/102). The HR-type HPV infection was observed in 12.6% (11/87) OSCC patients, out of which 45.5% (5/11) was HPV16-infected, 27.3% (3/11) had the HPV18 infection, while 27.3% (3/11) was infected by both HPV16 and HPV18. In this context, when we checked the distribution of total HR-HPV infection in our observed cancer sites, we found that buccal mucosa was the only HPV infected site and no tongue cancers had HPV infection. Apart from this, most of the OSCC specimens (>70%) were well-differentiated squamous cell carcinoma (WDSQCC) and majority were in advanced stage IV (54.9%), followed by stage III (18.6%).

Table 1.

Demographic characteristics of study subjects.

Variables	No. (%) of participants			Pearson’s chi-squarep-value
	Case 1	Case 2	Control
	(n = 50)	(n = 52)	(n = 74)
Age (years)
>49	35 (70.0)	39 (75.0)	20 (27.0)	<0.001
≤49	15 (30.0)	13 (25.0)	54 (73.0)	<0.001
Gender
Female	25 (50.0)	21 (40.4)	8 (10.8)	<0.001
Male	25 (50.0)	31 (59.6)	66 (89.2)	<0.001
HPV status (from PCR and IHC)
Total positive	4 (8.0)	7 (13.5)
HPV16	0 (0.0)	5 (9.6)
HPV18	2 (4.0)	1 (1.9)
HPV16 and HPV18	2 (4.0)	1 (1.9)
Total negative	36 (72.0)	40 (76.9)
Not determined	10 (20.0)	5 (9.6)
Pathology
MDSQCC	14 (28.0)	12 (23.1)
WDSQCC	36 (72.0)	40 (76.9)
Tumor staging
Stage I	1 (2.0)	2 (3.8)
Stage II	5 (10.0)	2 (3.8)
Stage III	14 (28.0)	5 (9.6)
Stage IV	24 (48.0)	32 (61.5)
Not determined	6 (12.0)	11 (21.2)
Specimen site
Tongue	7 (14.0)	12 (23.1)
Buccal mucosa	43 (86.0)	40 (76.9)

HPV: human papilloma virus; PCR: polymerase chain reaction; IHC: immunohistochemistry; MDSQCC: moderately differentiated squamous cell carcinoma; WDSQCC: well-differentiated squamous cell carcinoma.

p-value <0.05 is considered as statistically significant (bold); all of the p-values were estimated for case 1 and case 2 after comparing with the control group.

Identification of strongly associated SNPs from a stratified analysis: DFNA5 variant showed protective role, whereas SYK and NELL1 variants demonstrated harmful effect in OSCC

Our designed custom panel studied in the discovery set (n = 60) contained 1828 SNPs. After performing the targeted re-sequencing experiment, we found 115 SNPs significantly associated with OSCC (see Supplementary Table S3). We verified the same in confirmation set samples (n = 116) and finally obtained strongly associated SNPs are summarized in Supplementary Table S5. We have identified five SNPs significantly associated with tobacco-related OSCC in our studied population. Out of them, the variant rs2237306 from the DFNA5 region was found to have a protective role (OR = 0.33, p = 0.009) and the remaining four SNPs were positively associated (OR > 2.5, p < 0.05) with the disease. We also found that the identified SNPs present in SYK region were in complete linkage disequilibrium (D′= 1) and co-represented (r² > 0.8) (see Supplementary Figure S2).

Risk assessment of various external factors as well as the identified significant SNPs

While estimating the structure of our studied population, we observed that our heterogeneous samples from 12 different ethnic groups were clustered into the five major language families^50,51 as Indo-Aryan, Austro-Asiatic, Sino-Tibetan, Kuki-Chin, and Tibeto-Burman (see Supplementary Figure S3). Based on this knowledge, MLR analysis was performed to assess the risk of the factors considered in this study (Table 2).

Table 2.

Status of various external factors as well as SNPs and their association with OSCC.

Variables				Tobacco-related OSCC					Non-tobacco-related OSCC
				Case/ control	Adjusted OR	95% CI		p-value	Case/control	Adjusted OR	95% CI		p-value
						Lower	Upper				Lower	Upper
Age (years)	≤49			10/48	Reference				12/48	Reference
	>49			30/20	9.05	2.69	30.42	<0.001	35/20	7.00	3.03	16.18	<0.001
Gender	Male			21/60	Reference				28/60	Reference
Gender	Female			19/8	4.99	1.41	17.67	0.013	19/8	5.09	1.99	13.03	0.001
SNP details	Genotypes	Ref.	RA
rs2237306;^a 7p15.3;DFNA5	AA	C	C	6/14	Reference				6/14	Reference
	CA			7/20	3.48	0.63	19.19	0.153	19/20	2.22	0.71	6.96	0.173
	CC			27/34	3.61	0.98	13.37	0.054	22/34	1.51	0.50	4.52	0.461
	CA + CC			34/54	3.58	1.01	12.78	0.049	41/54	1.77	0.63	5.01	0.281
rs182361;9q22.2;SYK	CC	C	A	29/49	Reference				24/49	Reference
	CA			9/18	1.56	0.44	5.61	0.492	22/18	2.50	1.13	5.51	0.024
	AA			2/1	3.25	0.14	75.64	0.463	1/1	2.04	0.12	34.06	0.619
	CA + AA			11/19	1.71	0.50	5.79	0.391	23/19	2.47	1.13	5.39	0.023
rs290974;9q22.2;SYK	GG	G	A	29/49	Reference				30/49	Reference
	GA			8/17	1.52	0.41	5.69	0.534	15/17	1.44	0.63	3.30	0.388
	AA			3/2	2.73	0.24	30.65	0.417	2/2	1.63	0.22	12.21	0.633
	GA + AA			11/19	1.71	0.50	5.79	0.391	17/19	1.46	0.66	3.24	0.351
rs169724;9q22.2;SYK	TT	T	C	29/47	Reference				24/47	Reference
	CT			8/14	1.81	0.45	7.21	0.400	15/14	2.10	0.87	5.05	0.098
	CC			3/7	1.38	0.19	9.83	0.751	8/7	2.24	0.72	6.91	0.161
	CT + CC			11/21	1.67	0.49	5.68	0.411	23/19	2.14	0.99	4.63	0.052
rs1670661;11p15.1;NELL1	TT	T	C	20/57	Reference				31/57	Reference
	TC			19/11	6.33	1.67	24.02	0.007	14/11	2.34	0.95	5.77	0.065
	CC			1/0	Nd	Nd	Nd	Nd	2/0	Nd	Nd	Nd	Nd
	TC + CC			20/11	6.57	1.74	24.80	0.005	16/11	2.67	1.11	6.47	0.029

SNP: single nucleotide polymorphism; OSCC: oral squamous cell carcinoma; OR: odds ratio adjusted for population; CI: confidence interval; Ref.: reference allele; RA: risk allele; MLR: multinomial logistic regression; HPV: human papilloma virus; Nd: not determined.

This SNP demonstrated the protective effect. Unlike other SNPs, the minor/variant allele of this SNP is the protective allele; p-value <0.05 is considered as statistically significant (bold). For MLR analysis, we have considered 155 samples (case 1 = 40, case 2 = 47, and control = 68). The remaining samples (21 samples) were not considered for the analysis because in those samples, the HPV typing was unsuccessful (given in Table 1).

The subjects in the age group of >49 years had the 9.05-fold increased risk of tobacco-related OSCC and 7.00-fold increased risk of non-tobacco-related OSCC than those who are under the ≤49 years age group. Females had a 4.99-fold increased risk of the former OSCC type and 5.09-fold enhanced risk of latter as compared to males.

Apart from these physical factors, a significant 3.58-fold increased risk of tobacco-related OSCC was observed in the case of risk-allele-containing genotypes CA + CC of rs2237306 as compared to non-risk homozygous AA genotype. In addition, the risk-allele-containing genotypes TC + CC of rs1670661 was found to be associated with 6.57-fold and 2.67-fold enhanced risk of tobacco- and non-tobacco-related OSCC, respectively, and the heterozygous TC genotype was found to be associated with 6.33-fold increased risk of tobacco-related OSCC as compared to non-risk homozygous TT genotype. However, the risk-allele-containing genotypes CA + AA of rs182361 was found to be associated with 2.47-fold and heterozygous CA genotype was associated with 2.50-fold more risk of non-tobacco-related OSCC as compared to homozygous AA genotype.

MDR analysis to predict the multifactor models associated with both the types of OSCC

Interactions between gene–gene and gene–environmental factors for the occurrence of OSCC risk were evaluated by MDR analysis as summarized in Table 3. This analysis suggested that aging was the best “one-factor” model for both the types of OSCC (TBA = 0.7279, CVC = 100/100, p < 0.0001; TBA = 0.5299, CVC = 77/100, p < 0.0001, correspondingly) implying that the increasing age is the most significant factor for predicting the risk of OSCC. A combination of aging and rs1670661 variant was the best “two-factor” model (TBA = 0.6471, CVC = 88/100, p < 0.0001; TBA = 0.6067, CVC = 88/100, p < 0.0001, respectively) for both the types of OSCC. It suggests that this dual combination, identified across the 88 cross-validation subsets, involved in the causation of OSCC. The best “three-factor” model added the presence of HPV to the “two-factor” model for both the types of OSCC, but the aging got replaced by gender in the case of tobacco-related OSCC (TBA = 0.7581, CVC = 97/100, p < 0.0001; TBA = 0.7760, CVC = 100/100, p < 0.0001, respectively). Generally, the model with the highest TBA is taken as the best overall model. Therefore, the “three-factor” models were considered to be the best risk-predictive models for both the types of OSCC.

Table 3.

Summary of MDR analysis for OSCC risk prediction.

No. of locus	Best model	TrBA	TBA	CVC	p-value
Tobacco-related OSCC
First order	Age	0.7279	0.7279	100/100	<0.0001
Second order	Age, rs1670661	0.7721	0.6471	88/100	<0.0001
Third order^a	Gender, HPV, rs1670661	0.7906	0.7581	97/100	<0.0001
Fourth order	Age, gender, rs182361, rs1670661	0.7889	0.6971	91/100	<0.0001
Fifth order	Age, gender, rs182361, rs290974, rs1670661	0.7813	0.7522	96/100	<0.0001
Non-tobacco-related OSCC
First order	Age	0.7273	0.5299	77/100	<0.0001
Second order	Age, rs1670661	0.7346	0.6067	88/100	<0.0001
Third order^a	Age, HPV, rs1670661	0.8112	0.7760	100/100	<0.0001
Fourth order	Age, gender, HPV, rs1670661	0.7918	0.7498	84/100	<0.0001
Fifth order	Age, HPV, rs182361, rs290974, rs1670661	0.7535	0.6575	76/100	<0.0001

TrBA: training balance accuracy; TBA: testing balance accuracy; CVC: cross-validation consistency.

Best model predicted for oral cancer risk with the highest test balance and corresponding maximum CVC; p-value <0.05 is considered as statistically significant.

To determine whether the interactions were synergistic, non-synergistic, or redundant among the gene–gene and gene–environment factors under investigation, the entropy-based graphical model, that is, the “Interaction Entropy Graph” for OSCC-associated risk was constructed (Figure 1). In this graphical model, the age factor demonstrated the highest individual effect of 14.53% and SYK variant rs182361 demonstrated the lowest individual effect of 0.83% in tobacco-related OSCC (Figure 1(a)). Among the factors shown in the best risk-predictive “three-factor” model in this category, the presence of HPV (11.32%) showed a correlation with the NELL1 variant rs1670661 (9.84%) and gender (11.26%), whereas the effect of this variant was independent of gender. In the non-tobacco-related OSCC (Figure 1(b)), the presence of HPV showed the highest individual effect of 20.13%, and it showed the correlation with the NELL1 variant rs1670661 (4.44%) and aging (14.69%) in the best “three-factor” risk-predictive model. Similarly, the effect of this variant was independent of age and gender (7.94%). Interestingly, we observed that the SYK variants rs182361 and rs290974 were independent of age and correlated with gender in the case of tobacco-related OSCC. However, in non-tobacco-related OSCC, these two variants interacted with age and gender in an opposite manner. Moreover, all these variants in this analysis significantly correlated with the presence of HPV.

Figure 1.

Interaction entropy graph for showing the interactions among the various factors associated with (a) tobacco-related OSCC and (b) non-tobacco-related OSCC risk. Positive percentage of entropy indicates the synergistic interaction; negative values of entropy represent redundancy; golden lines suggest independence; and green and blue lines suggest redundancy or correlation.

Biological consequences of the variants derived by in silico approaches

According to the Ensembl VEP tool, all the variants identified from this study are intronic variants, especially the DFNA5 variant rs2237306, which is a nonsense-mediated mRNA decay (NMD) transcript variant (Transcript ID: ENST00000411476.2). The ESRsearch tool predicted that the G > T change (in reverse strand) in rs2237306 introduced a known ESR, hnRNP-B binding site. Similarly, T > C change in the NELL1 variant rs1670661 created a known ESR, SRp40 binding site and broke the potential exonic splicing silencer motif sequence. The G > T change in the SYK variant rs182361 introduced a potential exonic splicing silencer motif sequence. These results recommended a probable role of these variants in alternative splicing. The F-SNP tool used TFSEARCH v1.3 for the transcription factor binding site prediction. According to this tool, the C > A change in rs2237306 and T > C change in rs1670661 introduced the binding sites for the transcription factors, c-Myb and v-Myb, respectively. Apart from this, the enrichment map of the over-represented GO categories for SYK, DFNA5, and NELL1 genes was also generated (Figure 2) and in order to identify the biological significance of these genes, a gene–gene interaction network (Figure 3) was constructed.

Figure 2.

Involvement of the identified three genes in different biological pathways. Enrichment map of the over-represented Gene Ontology (GO) categories for SYK, DFNA5, and NELL1 genes has been generated from BiNGO (p-value < 0.001).

Figure 3.

Visualization of a co-expressed network. The figure represents the gene–gene network by GeneMANIA, where the significant genes interacting with our identified genes are demonstrated.

Discussion

The demographic characteristics of the studied NE Indian populations showed that the buccal mucosa was the most common (81.4%) cancer site, which is highly prevalent in India.⁵² Previously, it has been suggested that when betel quid usage is extensive, the buccal mucosa and tongue are the most frequent sites for OSCC.⁵³ We found that the majority of our patients (72.5%) in the age group of >49 years, as expected, as the incidence of oral cancer is age specific.³ The MLR analysis showed that the subjects older than 49 years were more susceptible to both the types of OSCC than those who were from the lower age group. MDR analysis suggested that the aging had the highest individual effect in case of tobacco-related OSCC. These results actually gave support for including only those individuals who had at least 10 years of chewing habits, as also suggested by earlier studies.^22,54 It implies that the prolonged tobacco-chewing habits increase the chances of being exposed to various carcinogenic compounds for a longer period.

Our observations are similar to a previous study,⁵⁵ suggesting females likely to be using smokeless (chewed) tobacco rather than smoking as compared to males in India. This has also been shown in a report from southern India.⁵⁶ Another study on North Indian population suggested that the use of SLT is a critical risk factor for females.⁵⁷ Furthermore, it was also reported that women are more susceptible to oral cancer than men in South Asian countries.⁵⁸ This study also showed that the females were ∼5-fold more likely to have both the types of OSCC than males as suggested by MLR analysis.

In this study, the HPV-infected site was ascertained in accordance with the previous studies^59,60 after taking the consensus observations from the PCR-based method as well as p16-immunohistochemistry study. Here, we observed 12.6% HR-HPV infected OSCC patients, which is lower as compared to the former studies on NE Indian population^8,61 but elevated than a report on a hospital-based study in North America.⁶² The evidence of HPV prevalence in oral cavity among Indian as well as global populations is emerging to be highly discordant. Studies with data based on developed nations accounted for a variable prevalence ranging from 10% to 80%.⁶³ While checking the global scenario on HPV prevalence, we found that the earlier population-based case-control studies in Sweden⁶⁴ and Mexico⁶⁵ reported a higher HPV prevalence (35% and 43.5%, respectively), whereas another retrospective study from Canada⁶⁶ showed a lesser HPV prevalence (19%). Similarly, in India, a study from the Southern region indicated no HPV infection,⁶⁷ whereas the prevalence of HPV in OSCC is reported to be different across the various regions with 6% in the western region,⁶⁸ 33.6% in the eastern region,⁶⁹ and 28%–43.5% in the northeastern region.^8,61 The higher HPV prevalence in NE India as compared to other regions of the Indian subcontinent is very likely due to the relatively higher consumption of tobacco products,¹⁷ as some earlier reports suggested that any form of tobacco use is probably linked to an increased risk of HPV infection.^70,71

Apart from the above-mentioned external factors, the identified genetic predisposing factors included the risk variants, rs290974 from SYK region and rs1670661 from NELL1 region, which are reported for the first time in this study. The other two SNPs (rs182361 and rs169724) from SYK region were previously reported in the GWAS conducted on western Indian population.²² The MDR analysis showed that the SYK variants were significantly correlated with the presence of HPV. The interactions among the presence of HPV, NELL1 variant rs1670661 along with age and gender increased the risk of both the types of OSCC.

The observations from the stratified analysis suggested that the genes like SYK and NELL1 may not be directly associated with the tobacco-related OSCC, but the newly identified variant in the DFNA5 region showed the protective role against tobacco-related OSCC. This observation was further verified by the MLR analysis, which reconfirmed that the variant rs1670661 from NELL1 region was associated with both the types of OSCC. However, the variant rs2237306 from DFNA5 region was associated with tobacco-related OSCC, while the variant rs182361 from SYK region had an association with non-tobacco-related OSCC. These make the DFNA5 as an important gene of interest for tobacco-related OSCC. The toxicogenomic analysis also suggested that Benzo(a)pyrene, one of the major chemical components of SLT products,^72,73 was involved in the increased expression of DFNA5 gene, the deafness-associated tumor suppressor gene (see Supplementary Figure S4). This chemical compound was previously known to be associated with miRNA-mediated activation of specific BaP-responsive pathways in human.⁷⁴ Interestingly, the DFNA5 gene, responsible for the autosomal dominant hearing loss⁷⁵ is inversely correlated with estrogen receptor expression in breast cancer.⁷⁶ The down-regulated DFNA5 contributes to acquired etoposide resistance in melanoma cells, as it is positively associated with etoposide-induced apoptotic pathway and caspase-3 activity.⁷⁷ Besides, this gene is positively involved in p53-induced apoptosis in response to DNA damage,⁷⁸ while mutDFNA5 (the mutated form of DFNA5)-induced cell death is mediated by the MAPK pathway, especially through the extracellular signal-regulated kinase (ERK) and the c-Jun N-terminal kinase (JNK) branch.⁷⁹

Results from different in silico tools suggested that the variants identified in this study were involved in the transcriptional regulation of the respective candidate genes. Interestingly, for rs2237306, we noted more than one transcripts of DFNA5 where this variant occurs. However, we have focused on the NMD transcript because we hypothesized that this particular NMD transcript may be predominant over other transcripts, activating the NMD pathway that can degrade other aberrant transcripts,⁸⁰ which has to be proved experimentally. Among the over-represented GO categories for SYK, DFNA5, and NELL1 genes, we observed that these genes were significantly involved in the regulation of GM-CSF biosynthetic process that has a role in immune system development by promoting defense against infections.⁸¹ Moreover, these genes were involved in positive regulation of IL-3 biosynthetic process, differentiation, and activation of gamma-delta T-cells that link innate and adaptive immune responses⁸² and also capable of phagocytosis.⁸³

From the pathway analysis, we infer that the DFNA5 gene that is co-expressed with the TERT gene physically interacts with SMG6 and SMG5, the NMD factors.⁸⁴ This DFNA5 gene is also involved in genetic interactions with the TEP1 and UBE3A genes.⁸⁵ However, the NELL1 gene is involved in genetic interactions with PTTG1 and SYK genes.⁸⁵ This PTTG1 gene that is co-expressed with the DFNA5 gene⁸⁶ physically interacts with TP53 gene and co-localized with SYK gene.⁸⁷ Moreover, this SYK gene that is co-localized with TP53 involved in genetic interactions with TEP1 and UBE3A genes,^85,87 which physically interact with TP53 gene. Interestingly, the UBE3A gene encodes for a viral pE6-associated protein (E6AP) and the E6/E6AP complex facilitates inactivation and degradation of the p53 protein as shown before.⁸⁸ Apart from this, in combination with the viral pE7 protein, HR-HPV pE6 protein is involved in the induction of telomerase activity by inducing hTERT expression at a transcriptional level. High telomerase activity, observed in more than 85% of cancer cells, strongly represents a crucial role in tumorigenesis.^89,90 However, further experimental verification of these observations is necessary.

The above-mentioned observations suggest that the NELL1 and SYK genes are probably involved in the HPV-mediated oncogenesis process and so, the SNPs on these genes represented potential risk for OSCC. However, DFNA5 gene variant has a possible protective role against tobacco-related OSCC by activating the NMD pathway, which degrades the aberrant DFNA5 transcript (Figure 4). This, in turn, activates the p53-induced apoptotic pathway in response to DNA damage.⁷⁸

Figure 4.

Proposed model for the probable mechanism of action of the identified protective SNP, rs2237306 in the DFNA5 genetic region.

In conclusion, our study approach can help in predicting the population-specific significant polymorphisms associated with OSCC in any heterogeneous population. On a larger scale, this would also help in developing an SNP panel for the early detection of the disease. Understanding the role of SNPs in different cell-signaling pathways may lead us to new therapeutic interventions.

Supplemental Material

Addional_file_1_revised_13-06-2018 – Supplemental material for Association of DFNA5, SYK, and NELL1 variants along with HPV infection in oral cancer among the prolonged tobacco-chewers

Supplemental material, Addional_file_1_revised_13-06-2018 for Association of DFNA5, SYK, and NELL1 variants along with HPV infection in oral cancer among the prolonged tobacco-chewers by Sharbadeb Kundu, Vijayalakshmi Ramshankar, Akalesh Kumar Verma, Soundara Viveka Thangaraj, Arvind Krishnamurthy, Rajeev Kumar, Ravi Kannan and Sankar Kumar Ghosh in Tumor Biology

Supplemental Material

Additional_file_2_revised_13-06-2018 – Supplemental material for Association of DFNA5, SYK, and NELL1 variants along with HPV infection in oral cancer among the prolonged tobacco-chewers

Supplemental material, Additional_file_2_revised_13-06-2018 for Association of DFNA5, SYK, and NELL1 variants along with HPV infection in oral cancer among the prolonged tobacco-chewers by Sharbadeb Kundu, Vijayalakshmi Ramshankar, Akalesh Kumar Verma, Soundara Viveka Thangaraj, Arvind Krishnamurthy, Rajeev Kumar, Ravi Kannan and Sankar Kumar Ghosh in Tumor Biology

Supplemental Material

Supplementary_Information_revised_13-06-2018 – Supplemental material for Association of DFNA5, SYK, and NELL1 variants along with HPV infection in oral cancer among the prolonged tobacco-chewers

Supplemental material, Supplementary_Information_revised_13-06-2018 for Association of DFNA5, SYK, and NELL1 variants along with HPV infection in oral cancer among the prolonged tobacco-chewers by Sharbadeb Kundu, Vijayalakshmi Ramshankar, Akalesh Kumar Verma, Soundara Viveka Thangaraj, Arvind Krishnamurthy, Rajeev Kumar, Ravi Kannan and Sankar Kumar Ghosh in Tumor Biology

Footnotes

Acknowledgements

The authors gratefully acknowledge the Department of Biotechnology (DBT), Govt. of India, for providing research grant; Department of Science & Technology (DST), Govt. of India, for providing fellowship to S.K. under DST-INSPIRE scheme (DST/INSPIRE Fellowship/2013/862). The authors are also thankful to Bioinformatics division of Invitrogen Bioservices India Pvt. Ltd. (Thermo Fisher Scientific) for their support in NGS data analysis and Mr Yogesh Pandey from Imperial Life Sciences for his support and help with Sequenom Mass Array Experimentation.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the NER Twinning project (BT/349/NE/TBP/2012), funded by the Department of Biotechnology (DBT), Govt. of India.

ORCID iD

Sankar Kumar Ghosh

Supplementary Information

Supplementary Information is available at Tumor Biology online.

References

Warnakulasuriya

. Global epidemiology of oral and oropharyngeal cancer. Oral Oncol 2009; 45: 309–316.

Krishnamurthy

Ramshankar

. Early stage oral tongue cancer among non-tobacco users–an increasing trend observed in a South Indian patient population presenting at a single centre. Asian Pac J Cancer Prev 2013; 14: 5061–5065.

Coelho

. Challenges of the oral cancer burden in India. J Cancer Epidemiol 2012; 2012: 701932.

Ferlay

Soerjomataram

Ervik

. GLOBOCAN, cancer incidence and mortality worldwide: IARC cancer base no. 11 [Internet]. Lyon: International Agency for Research on Cancer, 2013.

Varshitha

. Prevalence of oral cancer in India. J Pharm Sci Res 2015; 7: 845–848.

Gupta

Ariyawardana

Johnson

. Oral cancer in India continues in epidemic proportions: evidence base and policy initiatives. Int Dent J 2013; 63: 12–25.

Manoharan

Tyagi

Raina

. Cancer incidences in rural Delhi: 2004–05. Asian Pac J Cancer Prev 2010; 11: 73–77.

Mondal

Ghosh

Choudhury

et al . Mitochondrial DNA copy number and risk of oral cancer: a report from Northeast India. PLoS One 2013; 8: e57771.

Adams

Wise-Draper

Wells

. Human papillomavirus induced transformation in cervical and head and neck cancers. Cancer 2014; 6: 1793–1820.

10.

Gupta

Ray

. Smokeless tobacco and health in India and South Asia. Respirology 2003; 8: 419–431.

11.

Mouth Cancer Foundation, 2017, http://www.mouthcancerfoundation.org/patients-guide/chewing-tobacco

12.

Siddiqi

Shah

Abbas

et al . Global burden of disease due to smokeless tobacco consumption in adults: analysis of data from 113 countries. BMC Med 2015; 13: 194.

13.

Suliankatchi

Sinha

Rath

et al . Smokeless tobacco use is “replacing” the smoking epidemic in the South East Asia Region. Nicotine Tob Res. Epub ahead of print 22 December 2017. DOI: 10.1093/ntr/ntx272.

14.

Shield

Ferlay

Jemal

et al . The global incidence of lip, oral cavity, and pharyngeal cancers by subsite in 2012. CA Cancer J Clin 2017; 67: 51–64.

15.

Chaturvedi

Sarin

Seth

et al . India: steep decline in tobacco consumption in India reported in second Global Adult Tobacco Survey (GATS 2017). Tobacco Control, 2017, https://blogs.bmj.com/tc/2017/06/23/india-steep-decline-in-tobacco-consumption-in-india-reported-in-second-global-adult-tobacco-survey-gats-2017/

16.

National Cancer Registry Program, Indian Council of Medical Research. Three-year report of population based cancer registries 2012-2014: incidence, distribution, trends in incidence rates and projections of burden of cancer. Bangalore, India: National Cancer Registry Program, Indian Council of Medical Research

17.

Thakur

Prinja

Bhatnagar

et al . Widespread inequalities in smoking & smokeless tobacco consumption across wealth quintiles in States of India: need for targeted interventions. Indian J Med Res 2015; 141: 789–798.

18.

Bhattacharjee

Chakraborty

Purkaystha

. Prevalence of head and neck cancers in the north east-An institutional study. Indian J Otolaryngol Head Neck Surg 2006; 58: 15–19.

19.

Kreimer

Clifford

Boyle

et al . Human papillomavirus types in head and neck squamous cell carcinomas worldwide: a systematic review. Cancer Epidemiol Biomarkers Prev 2005; 14: 467–475.

20.

Gupta

. Role of human papillomavirus in oral squamous cell carcinoma and oral potentially malignant disorders: a review of the literature. Indian J Dent 2015; 6: 91–98.

21.

Ramírez-Pérez

. The role of human papillomavirus in oral squamous cell carcinoma. Plastic Aesth Res 2016; 3: 132–141.

22.

Bhatnagar

Dabholkar

Saranath

. Genome-wide disease association study in chewing tobacco associated oral cancers. Oral Oncol 2012; 48: 831–835.

23.

India Project Team of the International Cancer Genome Consortium. Mutational landscape of gingivo-buccal oral squamous cell carcinoma reveals new recurrently-mutated genes and molecular subgroups. Nat Commun 2013; 4: 2873.

24.

Pickering

Zhang

Yoo

et al . Integrative genomic characterization of oral squamous cell carcinoma identifies frequent somatic drivers. Cancer Discov 2013; 3: 770–781.

25.

Sambrook

Russell

. The Condensed Protocols from molecular cloning: a laboratory manual. Harbor, NY: Cold Spring Harbor Laboratory Press, 2006, p. 800.

26.

Ghosh

Choudhury

Hansa

et al . Human papillomavirus testing for suspected cervical cancer patients from Southern Assam by fast-PCR. Asian Pac J Cancer Prev 2011; 12: 749–751.

27.

Evans

Adamson

Simmons-Arnold

et al . Touchdown general primer (GP5+/GP6+) PCR and optimized sample DNA concentration support the sensitive detection of human papillomavirus. BMC Clin Pathol 2005; 5: 10.

28.

Gronhoj Larsen

Gyldenlove

Jensen

et al . Correlation between human papillomavirus and p16 overexpression in oropharyngeal tumours: a systematic review. Br J Cancer 2014; 110: 1587–1594.

29.

Chute

Aramouni

Brainard

et al . Hybrid Capture 2 human papilloma virus testing for head and neck cytology specimens. J Am Soc Cytopathol 2014; 3: 173–182.

30.

Barrett

Fry

Maller

et al . Haploview: analysis and visualization of LD and haplotype maps. Bioinformatics 2005; 21: 263–265.

31.

Purcell

Neale

Todd-Brown

et al . PLINK: a tool set for whole-genome association and population-based linkage analyses. Am J Hum Genet 2007; 81: 559–575.

32.

Pritchard

Stephens

Donnelly

. Inference of population structure using multilocus genotype data. Genetics 2000; 155: 945–959.

33.

Chhatre

Emerson

. StrAuto—automation and parallelization of structure analysis, 2016, http://strauto.popgen.org

34.

Evanno

Regnaut

Goudet

. Detecting the number of clusters of individuals using the software STRUCTURE: a simulation study. Mol Ecol 2005; 14: 2611–2620.

35.

Hahn

Ritchie

Moore

. Multifactor dimensionality reduction software for detecting gene–gene and gene–environment interactions. Bioinformatics 2003; 19: 376–382.

36.

Cattaert

Calle

Dudek

et al . Model-based multifactor dimensionality reduction for detecting epistasis in case-control data in the presence of noise. Ann Hum Genet 2011; 75: 78–89.

37.

Moore

Gilbert

Tsai

et al . A flexible computational framework for detecting, characterizing, and interpreting statistical patterns of epistasis in genetic studies of human disease susceptibility. J Theor Biol 2006; 241: 252–261.

38.

Choudhury

Singh

Kundu

et al . Tobacco carcinogen-metabolizing genes CYP1A1, GSTM1, and GSTT1 polymorphisms and their interaction with tobacco exposure influence the risk of head and neck cancer in Northeast Indian population. Tumour Biol 2015; 36: 5773–5783.

39.

Choudhury

Ghosh

. Gene-environment interaction and susceptibility in head and neck cancer patients and in their first-degree relatives: a study of Northeast Indian population. J Oral Pathol Med 2015; 44: 495–501.

40.

Singh

Choudhury

Kapfo

et al . Influence of the CYP1A1 T3801C polymorphism on tobacco and alcohol-associated head and neck cancer susceptibility in Northeast India. Asian Pac J Cancer Prev 2015; 16: 6953–6961.

41.

Manuguerra

Matullo

Veglia

et al . Multi-factor dimensionality reduction applied to a large prospective investigation on gene–gene and gene–environment interactions. Carcinogenesis 2006; 28: 414–422.

42.

McLaren

Gil

Hunt

et al . The Ensembl Variant Effect Predictor. Genome Biol 2016; 17: 122.

43.

Goren

Ram

Amit

et al . Comparative analysis identifies exonic splicing regulatory sequences–the complex definition of enhancers and silencers. Mol Cell 2006; 22: 769–781.

44.

Lee

Shatkay

. F-SNP: computationally predicted functional SNPs for disease association studies. Nucleic Acids Res 2008; 36: D820–184.

45.

Davis

Grondin

Johnson

et al . The comparative toxicogenomics database: update 2017. Nucleic Acids Res 2017; 45: D972–19D8.

46.

Maere

Heymans

Kuiper

. BiNGO: a Cytoscape plugin to assess overrepresentation of gene ontology categories in biological networks. Bioinformatics 2005; 21: 3448–3449.

47.

Merico

Isserlin

Stueker

et al . Enrichment map: a network-based method for gene-set enrichment visualization and interpretation. PLoS One 2010; 5: e13984.

48.

Feng

Stein

. A human functional protein interaction network and its application to cancer data analysis. Genome Biol 2010; 11: R53.

49.

Mostafavi

Ray

Warde-Farley

et al . GeneMANIA: a real-time multiple association network integration algorithm for predicting gene function. Genome Biol 2008; 9(Suppl. 1): S4.

50.

Kumar

. Language Scenario in the North-east. Ishani 2008; 2: 1–7.

51.

Moral

. North-East India as a linguistic area. Monkhmer Studies 1997; 27: 43–54.

52.

Gupta

Johnson

. Systematic review and meta-analysis of association of smokeless tobacco and of betel quid without tobacco with incidence of oral cancer in South Asia and the Pacific. PLoS One 2014; 9: e113385.

53.

Chi

Day

Neville

. Oral cavity and oropharyngeal squamous cell carcinoma—an update. CA Cancer J Clin 2015; 65: 401–421.

54.

Zhou

Michaud

Langevin

et al . Smokeless tobacco and risk of head and neck cancer: evidence from a case-control study in New England. Int J Cancer 2013; 132: 1911–1917.

55.

American Cancer Society. Tobacco use in India: an evil with many faces (no. 011286-Rev.06/10 India Project). Cancer Patients Aid Association, Total Management of Cancer, American Cancer Society, 2009, http://indiancancersocietydelhi.in/wp-content/uploads/2016/04/Tobacco_Abuse_India.pdf

56.

Balaram

Sridhar

Rajkumar

et al . Oral cancer in southern India: the influence of smoking, drinking, paan-chewing and oral hygiene. Int J Cancer 2002; 98: 440–445.

57.

Singh

Kumar

Agarwal

et al . Clinico-epidemiological study of oral squamous cell carcinoma: a tertiary care centre study in North India. J Oral Biol Craniofac Res 2016; 6: 31–34.

58.

Niaz

Maqbool

Khan

et al . Smokeless tobacco (paan and gutkha) consumption, prevalence, and contribution to oral cancer. Epidemiol Health 2017; 39: e2017009.

59.

Mravak-Stipetic

Sabol

Kranjcic

et al . Human papillomavirus in the lesions of the oral mucosa according to topography. PLoS One 2013; 8: e69736.

60.

Sanchez-Vargas

Diaz-Hernandez

Martinez-Martinez

. Detection of human papillomavirus (HPV) in oral mucosa of women with cervical lesions and their relation to oral sex practices. Infect Agent Cancer 2010; 5: 25.

61.

Kumar

Rai

Das

et al . Alcohol and tobacco increases risk of high risk HPV infection in head and neck cancer patients: study from North-East Region of India. PLoS One 2015; 10: e0140700.

62.

Lingen

Xiao

Schmitt

et al . Low etiologic fraction for high-risk human papillomavirus in oral cavity squamous cell carcinomas. Oral Oncol 2013; 49: 1–8.

63.

Kumaraswamy

Vidhya

. Human papilloma virus and oral infections: an update. J Cancer Res Ther 2011; 7: 120–127.

64.

Hansson

Rosenquist

Antonsson

et al . Strong association between infection with human papillomavirus and oral and oropharyngeal squamous cell carcinoma: a population-based case-control study in southern Sweden. Acta Otolaryngol 2005; 125: 1337–1344.

65.

Anaya-Saavedra

Ramirez-Amador

Irigoyen-Camacho

et al . High association of human papillomavirus infection with oral cancer: a case-control study. Arch Med Res 2008; 39: 189–197.

66.

Pintos

Black

Sadeghi

et al . Human papillomavirus infection and oral cancer: a case-control study in Montreal, Canada. Oral Oncol 2008; 44: 242–250.

67.

Laprise

Madathil

Allison

et al . No role for human papillomavirus infection in oral cancers in a region in southern India. Int J Cancer 2016; 138: 912–917.

68.

Koppikar

deVilliers

Mulherkar

. Identification of human papillomaviruses in tumors of the oral cavity in an Indian community. Int J Cancer 2005; 113: 946–950.

69.

Nagpal

Patnaik

Das

. Prevalence of high-risk human papilloma virus types and its association with P53 codon 72 polymorphism in tobacco addicted oral squamous cell carcinoma (OSCC) patients of Eastern India. Int J Cancer 2002; 97: 649–653.

70.

Chakrobarty

Roy

Majumdar

et al . Relationship among tobacco habits, human papilloma virus (HPV) infection, p53 polymorphism/mutation and the risk of oral squamous cell carcinoma. J Oral Maxillofac Pathol 2014; 18: 211–216.

71.

Fakhry

Gillison

D’Souza

. Tobacco use and oral HPV-16 infection. JAMA 2014; 312: 1465–1467.

72.

Borgerding

Bodnar

Curtin

et al . The chemical composition of smokeless tobacco: a survey of products sold in the United States in 2006 and 2007. Regul Toxicol Pharmacol 2012; 64: 367–387.

73.

Sangar

. Smokeless tobacco: chemical composition and carcinogenicity. Int Res J Interdisciplinary & Multidisciplinary Stud 2016; 2: 7–16.

74.

Lizarraga

Gaj

Brauers

et al . Benzo[a]pyrene-induced changes in microRNA-mRNA networks. Chem Res Toxicol 2012; 25: 838–849.

75.

De Beeck

Van Laer

Van Camp

. DFNA5, a gene involved in hearing loss and cancer: a review. Ann Otol Rhinol Laryngol 2012; 121: 197–207.

76.

Thompson

Weigel

. Characterization of a gene that is inversely correlated with estrogen receptor expression (ICERE-1) in breast carcinomas. Eur J Biochem 1998; 252: 169–177.

77.

Lage

Helmbach

Grottke

et al . DFNA5 (ICERE-1) contributes to acquired etoposide resistance in melanoma cells. FEBS Lett 2001; 494: 54–59.

78.

Masuda

Futamura

Kamino

et al . The potential role of DFNA5, a hearing impairment gene, in p53-mediated cellular response to DNA damage. J Hum Genet 2006; 51: 652–664.

79.

Van Rossom

Op de Beeck

Hristovska

et al . The deafness gene DFNA5 induces programmed cell death through mitochondria and MAPK-related pathways. Front Cell Neurosci 2015; 9: 231.

80.

Baker

Parker

. Nonsense-mediated mRNA decay: terminating erroneous gene expression. Curr Opin Cell Biol 2004; 16: 293–299.

81.

Subramanian Vignesh

Landero Figueroa

Porollo

et al . Granulocyte macrophage-colony stimulating factor induced Zn sequestration enhances macrophage superoxide and limits intracellular pathogen survival. Immunity 2013; 39: 697–710.

82.

Holtmeier

Kabelitz

. Gammadelta T cells link innate and adaptive immune responses. Chem Immunol Allergy 2005; 86: 151–183.

83.

Wong

et al . Human gamma delta T cells: a lymphoid lineage cell capable of professional phagocytosis. J Immunol 2009; 183: 5622–5629.

84.

Maquat

. Nonsense-mediated mRNA decay: splicing, translation and mRNP dynamics. Nat Rev Mol Cell Biol 2004; 5: 89–99.

85.

Lin

Wang

Ahn

et al . A genome-wide map of human genetic interactions inferred from radiation hybrid genotypes. Genome Res 2010; 20: 1122–1132.

86.

Cheok

Yang

Pui

et al . Treatment-specific changes in gene expression discriminate in vivo drug response in human leukemia cells. Nat Genet 2003; 34: 85–90.

87.

Johnson

Castle

Garrett-Engele

et al . Genome-wide survey of human alternative pre-mRNA splicing with exon junction microarrays. Science 2003; 302: 2141–2144.

88.

Martinez-Zapien

Ruiz

Poirson

et al . Structure of the E6/E6AP/p53 complex required for HPV-mediated degradation of p53. Nature 2016; 529: 541–545.

89.

Munger

Baldwin

Edwards

et al . Mechanisms of human papillomavirus-induced oncogenesis. J Virol 2004; 78: 11451–11460.

90.

Narisawa-Saito

Kiyono

. Basic mechanisms of high-risk human papillomavirus-induced carcinogenesis: roles of E6 and E7 proteins. Cancer Sci 2007; 98: 1505–1511.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.57 MB

0.85 MB

0.58 MB

Association of DFNA5,SYK,and NELL1 variants along with HPV infection in oral cancer among the prolonged tobacco-chewers

Abstract

Keywords

Introduction

Materials and methods

Sampling and demographic data collection

Genomic DNA extraction

Targeted re-sequencing in the Ion PGMTM platform on the discovery set

Validation and replication in Sequenom iPLEX MassARRAYTM platform on the confirmation set

PCR-based HPV genotyping

Immunohistochemistry for p16 expression

Statistical analysis

MDR approach

Interaction entropy graph

In silico analysis

Results

Demographic characteristics of the participants: females are more susceptible to tobacco-related OSCC than males

Identification of strongly associated SNPs from a stratified analysis: DFNA5 variant showed protective role, whereas SYK and NELL1 variants demonstrated harmful effect in OSCC

Risk assessment of various external factors as well as the identified significant SNPs

MDR analysis to predict the multifactor models associated with both the types of OSCC

Biological consequences of the variants derived by in silico approaches

Discussion

Supplemental Material

Addional_file_1_revised_13-06-2018 – Supplemental material for Association of DFNA5, SYK, and NELL1 variants along with HPV infection in oral cancer among the prolonged tobacco-chewers

Supplemental Material

Additional_file_2_revised_13-06-2018 – Supplemental material for Association of DFNA5, SYK, and NELL1 variants along with HPV infection in oral cancer among the prolonged tobacco-chewers

Supplemental Material

Supplementary_Information_revised_13-06-2018 – Supplemental material for Association of DFNA5, SYK, and NELL1 variants along with HPV infection in oral cancer among the prolonged tobacco-chewers

Footnotes

Acknowledgements

Declaration of conflicting interests

Funding

ORCID iD

Supplementary Information

References

Supplementary Material

Targeted re-sequencing in the Ion PGM^TM platform on the discovery set

Validation and replication in Sequenom iPLEX MassARRAY^TM platform on the confirmation set